Abstract WP124: Risk Factors for Ischemic Events in Patients with Embolic Stroke of Undetermined Source: A Post-Hoc Analysis of the ARCADIA Trial

Introduction: The risk of recurrent ischemic events in patients with cryptogenic stroke is nearly 5% per year. Data is limited regarding factors associated with recurrence risk and whether such features could identify subgroups who may benefit from anticoagulation. We used the ARCADIA trial data to identify risk factors of recurrent ischemic stroke and whether anticoagulation is beneficial in subgroups defined by these risk factors. Methods: In the ARCADIA trial, patients with cryptogenic stroke and evidence of atrial cardiopathy based on biomarkers were randomized to aspirin or apixaban. Those with any history of atrial fibrillation or a left ventricular ejection fraction (LVEF) less than 30% were excluded. In this post-hoc analysis, the primary outcome for this analysis was recurrent ischemic stroke or systemic embolism. We used univariate analyses to identify demographic, clinical, laboratory, and imaging markers associated with recurrence risk (p < 0.05) and built a Multivariable Cox regression model including variables achieving statistical significance on univariate analyses to identify risk factors of the primary outcome. We then compared outcome rates between apixaban and aspirin in patients with each of these high-risk features. Results: Of 1015 patients included in the ARCADIA trial, 77 (7.6%) had a recurrent ischemic stroke or systemic embolism. In Multivariable Cox regression analysis, factors associated with recurrent ischemic stroke or systemic embolism were history of diabetes (aHR 2.37 95% CI 1.50-3.83), prior stroke or TIA (aHR 2.08 95% CI 1.25-3.47), and LVEF < 55% (aHR 1.81 95% CI 1.03-3.18). Only patients with a reduced LVEF (<55%) (~14% of patients) seemed to benefit from apixaban over aspirin (aHR 0.54 95% CI 0.30-0.96). Conclusions: In this post-hoc analysis of the ARCADIA trial, history of diabetes, prior stroke or TIA, and reduced ejection fraction were associated with increased recurrence risk, but only those with reduced ejection fraction seemed to benefit from anticoagulation.